Alzheimer disease (AD) is a devastating neurodegenerative disorder that affects millions of individuals in the U.S. It has so far resisted attempts to develop effective therapies despite numerous (failed) clinical trials based on known targets, most identified over 20 years ago. While genomic research (e.g. the Alzheimer’s Disease Sequencing Project; ADSP) has identified numerous additional risk loci, these results derive primarily from case- control datasets. In contrast, cohorts designed to identify variants that may protect from AD, and those using complementary study designs, are few. We used our extensive experience with the Amish communities in Indiana and Ohio to establish a cohort of older individuals at high risk of developing AD but who are cognitively unimpaired (CU). The Amish provide a unique opportunity to identify protective variants for AD because of their reduced background genetic variation and environmental risk factors. Their small founding population and endogamy provides enrichment for rare variants. Founder populations also enable testing for non-additive allelic effects and can magnify sub-significant association signals identified in case-control studies. The stability and engagement of our Amish participants enables longitudinal assessments of cognition and biomarkers. Our primary goals are to identify AD protective loci and characterize pre-clinical biomarkers of progression to cognitive impairment. Building on our existing large cohort, our replicated protective locus and several suggestive protective loci, and our existing biobank of DNA and plasma and databank of GWAS and WGS, we propose to: 1) Perform longitudinal assessment of cognition in our family-based Amish cohort; 2) Identify protective factors for AD and predictors of progression to cognitive impairment by analyzing genomic and longitudinal cognitive, biomarker, and SDOH data; and 3) Examine the functional implications of current and novel genes and variants by prioritizing loci using in silico annotation for functional consequences followed by in vitro functional characterization. Our results will identify potential druggable targets and accelerate the development of better treatments for AD.